Maintenance and eventual replacement are inevitable elements of modern infrastructure. Assets break, so it behooves us to monitor them and try to predict and catch problems early.
But what do you do if the asset in question is telecom infrastructure — a fiber-optic cable? These glass filaments don’t age in the same way other materials do. That makes it exponentially harder to conduct capital planning — how do you know where to start?
That’s exactly what we’ve been tackling with some cutting-edge asset health research with Duke Energy and Southern Company. The objectives are to monitor and study fiber performance, observe patterns, and ultimately deploy our findings to reduce downtime and repair expenses.
We have installed monitoring units on several fiber-optic lines to study ongoing performance and efficiency. By compiling a record of their historical and ongoing line performance, we are amassing the data necessary to build predictive models of the telecommunications system’s behavior. This data is helping us develop analytic algorithms for diagnosing potential issues, which subsequently assists with developing efficient maintenance strategies.
Fiber-optic cable monitoring involves continuous automated surveillance, optical time domain reflectometer (OTDR) trace comparison, automatic alarms, network documentation, reporting and trend analysis. More simply put, we test the light transmission rate over the cable continuously, observing and documenting changes over time. The data can help us examine a number of questions:
- How much loss of transmission rate has the fiber experienced since installation?
- Has there been a cable break?
- How quickly does deterioration occur?
- Does the algorithm sufficiently predict real-world findings as it is refined?
The research benefited from a break in December — literally. One of the fiber-optic cables we were monitoring experienced a physical failure. While this was unfortunate for the utility, it gave us a significant data point for studying the asset health records and refining our algorithms.
We will continue the field research and refining process, because the potential payoff for utilities is significant. The more insight that can be shed on asset health, the better prepared those assets’ owners will be to optimize maintenance and minimize outages and time-consuming repair or replacement projects.
For more insights on the latest in power planning from key industry thinkers, check out Amplified Perspectives.
